Thirty-nine percent of reproductive age women have used a non-steroidal anti-inflammatory drug (NSAID) in the past week, and NSAIDs are the most frequent medication exposure in the first trimester of pregnancy. The safety of NSAID use in early pregnancy is uncertain. Our objective is to examine prospectively the association of NSAID use with spontaneous abortion. We propose to enroll women who are planning a pregnancy and obtain daily diaries of medication use, symptoms (indications) and other key confounders during the period while participants are attempting to conceive. Among the first 1,200 women who become pregnant, we will continue to collect daily diary information through the 12th week of pregnancy, obtain an early pregnancy ultrasound in the sixth week, and conduct an extensive computer assisted telephone interview at 13 weeks. Birth outcomes will be documented for all pregnancies and karyotyping will be done on all losses where a tissue specimen is available. These methods will allow us to answer our primary research questions: (1) Is NSAID use independently associated with increased risk of spontaneous abortion? (2) Does the association of NSAIDs with spontaneous abortion depend on cumulative dose, regular versus episodic use, and/or exposure during critical developmental windows? and these secondary questions: (3) Are common indications for NSAID use independently associated with spontaneous abortion? (4) Does developmental stage attained prior to spontaneous abortion differ between exposed and unexposed pregnancies? The primary tools for the proposed study are already in place including successful approaches to community-based recruitment prior to pregnancy, ultrasound forms and protocols, computer assisted telephone interviews, data management and quality control systems and experienced staff. Costly piloting and implementation phases have been completed. In short, we are fully prepared to efficiently address pressing questions about NSAID exposure in early pregnancy by conducting a large-scale, fully-prospective cohort study. If NSAID use has an adverse effect on early pregnancy outcomes, the public health impact would be substantial. For example, if NSAID exposure increases relative risk by 50 per cent, and 19 per cent of women use NSAIDs during the first trimester of pregnancy (a conservative estimate based on retrospective data), then over 80,000 spontaneous abortions could be attributed to NSAIDs each year in the United States.